5. banking sector performance
TRANSCRIPT
Half-YearlY MonetarY and financial StabilitY report SepteMber 201360
5. banking sector performance
Notwithstanding an uncertain external environment and the slowdown of the Mainland
China economy, the Hong Kong banking sector continued to post vibrant earnings
performance. While bank lending grew more rapidly in the first half of 2013, domestic
liquidity conditions remained healthy. However, the banking sector will face rising risks in
the year to come. Domestically, the risk of interest rate volatility is expected to escalate. In
the process of the eventual normalisation of US monetary policy, the impact of any shift of
the yield curve or significant rise in interest rates on banks’ balance sheets and asset quality
may not be small. In particular, corporations’ debt servicing ability would be under test given
rising corporate leverage, and property prices could be under significant pressures. Externally,
the rising share of banks’ Mainland exposure continues to be a significant risk factor. The
need for continued stringent prudential management of such exposure cannot be
over-emphasised.
5.1 Profitability and capitalisation
ProfitabilityThe banking sector continued to post vibrant
performance, with the aggregate pre-tax
operating profits of retail banks 31 growing by
30.9% in the first half of 2013 from the second
half of 2012. Their return on assets 32 also rose to
1.39%, compared with a return of 1.1% in the
previous six months (Chart 5.1). Improved net
interest and non-interest incomes, which grew
by 7.8% and 19.6% respectively, plus better
control of operating expenses and lower net
provision, were the main drivers of profitability
growth.
Chart 5.1Profitability of retail banks
% of total assets% of total assets
Note: Semi-annually annualised figures.
Source: HKMA.
H1 H2 H1 H2 H1 H2 H1 H2 H1 H1H2 H1 H2 H1H22006 2007 2008 2009 20112010 20132012
Non-interest income (lhs)Net interest income (lhs)General and administrative expenses (lhs)Net provision (lhs)Pre-tax operating profit (rhs)
-2
-1
0
1
2
3
4
0.50
0.75
1.00
1.25
1.50
1.75
2.00
31 Throughout this chapter, figures for the banking sector relate to Hong Kong offices only, except where otherwise stated.
32 Return on assets is calculated based on aggregate pre-tax operating profits.
61 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
The net interest margin of retail banks improved
further in the first half of 2013 to 1.41%, from
1.37% in the second half of 2012 (Chart 5.2),
partly due to an easing of banks’ funding
pressure. The composite interest rate, a measure
of the average cost of Hong Kong dollar funds of
retail banks, averaged 0.27% in the first half,
down from 0.37% in the second half of 2012
(Chart 5.3).
%
Chart 5.2 Net interest margin of retail banks
Note: Quarterly annualised figures.
Source: HKMA.
2006 2007 2008 2009 2010 2013(Q1 - Q2)
201220111.1
1.3
1.5
1.7
1.9
2.1
%
Chart 5.3 Interest rates
Notes:(a) End of period figures.(b) Period-average figures for approved loans. All mortgage rates are estimates only.
Sources: HKMA and staff estimates.
2006 2007 2008 2009 2010 2013(Jan - Jul)
20122011
Composite interest rate (a)3-month HIBOR (a)Average mortgage rate for BLR-based mortgages (b)Average mortgage rate for HIBOR-based mortgages (b)
0
1
2
3
4
5
6
An analysis of licensed banks’ data shows that
their overall interest costs have fallen, with the
market-based funding cost down by 4 basis
points and the deposit funding cost remaining
largely stable (Chart 5.4). 33
Chart 5.4Hong Kong and US dollar funding cost and maturity of licensed banks
Number of months
0.0
3.5
Dec 2010 Jun 2011 Dec 2011 Jun 2012 Dec 2012 Jun 2013
Source: HKMA.
Average Maturity (rhs)Market based funding cost (lhs)
%
0.0
1.2
0.2
0.4
0.6
0.8
1.0
0.5
1.0
1.5
2.0
2.5
3.0
Funding cost of deposits (lhs)Overall funding cost (lhs)
Mortgage rates on average remained low in the
first quarter of the year, with the best lending
rate-based (BLR-based) rate edging higher but the
HIBOR-based rate softening marginally. Partly in
response to the rise in BLR-based rate, the share
of BLR-based mortgages amongst newly approved
mortgage loans decreased to 82.4% in the first
half of 2013, from 92.8% in the second half of
2012.
33 Market-based funding cost is measured by the interest costs of banks’ non-deposit interest bearing liabilities.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201362
banking sector performance
CapitalisationCapitalisation of the banking sector remained
well above the minimum international
standards. The consolidated capital adequacy
ratio of locally incorporated AIs edged up to
15.9% at the end of June, from 15.7% at the end
of 2012 (Chart 5.5), with the tier-one capital
adequacy ratio (the ratio of tier-one capital to
total risk-weighted assets) edging lower to 13.2%,
from 13.3%.
%
Chart 5.5 Capitalisation of locally incorporated AIs
8
12
16
14
10
18
20
2007 2008 2009 2010 2013(Q1 - Q2)
20122011
Notes: 1. Consolidated positions. 2. With effect from 1 January 2013, a revised capital adequacy framework (Basel III)
was introduced for locally incorporated AIs. The capital adequacy ratios from March 2013 onwards are therefore not directly comparable with those up to December 2012.
Source: HKMA.
Tier-one capital adequacy
Capital adequacy ratio
Introduction of Basel III
5.2 Liquidity and funding
Domestic liquidity conditions remained healthy.
Although the average liquidity ratio of all AIs fell
moderately in the first half of 2013, it remained
well above the regulatory minimum of 25%. For
retail banks, the ratio decreased to 38.9% at the
end of June, from 42.6% at the end of 2012
(Chart 5.6). The fall partly reflected banks’ recent
moves of reallocating assets from more liquid
instruments (such as interbank claims) into less
liquid assets (such as customer loans and
advances).
%
Chart 5.6 Liquidity ratio of retail banks
Note: Quarterly average figures.
Source: HKMA.
30
40
45
35
50
55
2006 2007 2008 2009 2010 2013(Q1 - Q2)
20122011
Customer deposits continued to be the primary
funding source for retail banks, although their
share to banks’ total liabilities decreased
marginally to 73.9% at the end of June, from
74.2% at the end of 2012 (Chart 5.7).
Chart 5.7 Liabilities structure of retail banks
Notes:1. Figures may not add up to total due to rounding.2. Figures refer to the percentage of total liabilities (including capital and reserves). 3. Debt securities comprise negotiable certificates of deposit and all other
negotiable debt instruments.
Source: HKMA.
Due to AIs Due to banks abroadCustomer deposits
Debt securities outstandingOther liabilities and capital and reserves
December 2012
74.2%
4.7%2.7%
14.5%
3.9%
5.2%3.1%
14.9%
2.8%
73.9%June 2013
63 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
34 For details, please refer to Section 4.2 of the report.
Due to a much stronger credit demand in the
first half of 2013, which outpaced the moderate
growth of deposits, the all currency loan-to-
deposit (LTD) ratio of all AIs rose to 71.9% at the
end of June, from 67.1% six months earlier
(Chart 5.8). In particular, the foreign currency
LTD ratio rose sharply by 6.2 percentage points
to 60.4%, which was mostly driven by a
significant increase in foreign currency loans and
trade financing. 34
%
Chart 5.8 Loan-to-deposit ratios of all AIs
Source: HKMA.
0
40
60
20
80
100
2006 2007 2008 2009 2010 2013(Q1-Q2)
20122011
Hong Kong dollar All currencies Foreign currencies
For retail banks, the all currency LTD ratio rose to
58.5% from 54.8%, with both the Hong Kong
dollar and foreign currency LTD ratios increasing
notably (Chart 5.9).
Chart 5.9Loan-to-deposit ratios of retail banks
Source: HKMA.
Foreign currenciesHong Kong dollar All currencies
0
100%
2006 2007 2008 2009 20122011 2013(Q1-Q2)
2010
20
40
60
80
Foreign currency positionThe banking sector’s capability to repay liabilities
denominated in foreign currencies can be
assessed by reference to the aggregate net open
position of AIs for all foreign currencies. This
position amounted to HK$69 billion at the end
of June 2013, which was equivalent to 0.4% of
total assets of AIs, indicating that the overall
exposure of AIs to foreign exchange risks may
not be of significant concern.
5.3 Interest rate risk
The spreads between the long- and short-term
interest rates for the US dollar and Hong Kong
dollar continued their upward trends, widening
further to over 200 basis points in August 2013
(Chart 5.10). This suggests that the incentive for
banks to search for yield by borrowing short-term
funds to purchase long-term interest-bearing
assets may have increased. This could potentially
lead to greater maturity mismatches and
increased interest rate risk.
Basis points
Chart 5.10Term spreads of Hong Kong and US dollars
Note: Term spreads are defined as 10-year swap rates minus three-month money market rates of the respective currencies.
Source: HKMA staff estimates based on data from Bloomberg.
-200
0
100
300
200
-100
400
500
2006 2007 2008 2009 2010 2013(Jan - Aug)
2011 2012
US dollar term spread Hong Kong dollar term spread
Half-YearlY MonetarY and financial StabilitY report SepteMber 201364
banking sector performance
Such interest rate risk should not be underrated,
as the possible impact of any significant shift in
the yield curve or rise in interest rates on banks’
balance sheets in the process of the eventual
normalisation of US monetary policy could be
severe. It is estimated that under a hypothetical
shock of an across-the-broad 200-basis-point
increase in interest rates, the economic value of
retail banks’ interest rate positions could be
subject to a decline equivalent to 1.4% of their
total capital base as of June 2013 (Chart 5.11).
While the impact appears to be manageable 35, a
significantly larger interest rate hike or an
unfavourable change in the shape of the yield
curve could result in a much bigger impact.
Chart 5.11Impact of interest rate shock on retail banks
% of total capital base
2008 2009 2010 20122011 2013(Q1 - Q2)
Notes:1. Interest rate shock refers to a standardised 200-basis-point parallel rate shock to
institutions’ interest rate risk exposures.2. The impact of the interest rate shock refers to its impact on the economic value of
banking and trading book 36, expressed as a percentage of the total capital base of banks.
Source: HKMA staff estimates.
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0
1.6
5.4 Credit risk
The asset quality of retail banks’ loan portfolios
continued to improve in the first half of the year,
with the classified loan ratio falling further to
0.45% at the end of June, from 0.48% six months
earlier. The ratio of overdue and rescheduled
loans also edged down to 0.34% from 0.39%
(Chart 5.12) during the period.
% of total loans
Chart 5.12Asset quality of retail banks
0
2
3
6
5
4
1
7
8
2001 20062005200420032002 2007 2008 2009 2010 2013(Q1 - Q2)
20122011
Notes: 1. Classified loans are those loans graded as “sub-standard”, “doubtful” or “loss”.2. Figures related to retail banks’ Hong Kong office(s) and overseas branches.
Source: HKMA.
Classified loans (gross) Overdue and rescheduled loans
Loan growth in the banking sector accelerated in
the first half of 2013. Total domestic lending 37 of
AIs increased by 10.2% in the first half, following
an increase of 4.1% in the preceding six months.
The rapid credit expansion was mainly
contributed by trade finance and loans to
corporations, whereas property-related loans
registered a more moderate increase of 3.2%,
after rising by 4.3% in the second half of 2012.
35 The HKMA will be particularly attentive to the capital sufficiency of “outlier AIs” – those whose interest rate risk leads to an economic value decline of more than 20% of their capital base as a result of applying the standardised interest rate shock to the banking book. For details, see HKMA Supervisory Policy Manual’s module “Interest Rate Risk Management” issued in December 2002.
36 Locally incorporated AIs subject to the market risk capital adequacy regime are required to report positions in the banking book only. Other locally incorporated AIs exempted from the market risk capital adequacy regime and overseas incorporated institutions are required to report aggregate positions in the banking book and trading book.
37 Defined as loans for use in Hong Kong plus trade-financing loans. If other loans for use outside Hong Kong are included, the growth in bank lending in the six months ended June 2013 is 9.5%, compared with 4.7% in the preceding six months.
65 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Looking ahead, the rapid growth in lending may
moderate. As suggested by results of the HKMA
Opinion Survey on Credit Condition Outlook of
June 2013, while more than two-thirds of the
respondents expected loan demand to remain at
the current level, the share of surveyed AIs
expecting lower loan demand in the next three
months had exceeded those expecting higher
loan demand (Table 5.A).
Table 5.AExpectation of loan demand in the next three months
(% of total respondents) Sep 2012 dec 2012 Mar 2013 Jun 2013
Considerably higher 0 0 0 0Somewhat higher 10 14 10 10Same 71 76 67 71Somewhat lower 19 10 24 19Considerably lower 0 0 0 0
Total 100 100 100 100
Note: Figures may not add up to 100% due to rounding.
Source: HKMA.
Household exposureHousehold loans 38 showed signs of moderation,
with the half-yearly growth rate slowing down to
3.8%, from 6.5% in the second half of 2012.
Mortgage lending, which is the major
component of household loans, grew at a slightly
slower pace of 3.1%, after expanding by 5% in
the second half of 2012, partly reflecting
softened property prices and reduced transaction
volumes. As a result, the share of mortgage
lending to total domestic loans edged down to
21.5% (Table 5.B).
Table 5.BHalf-yearly growth of loans to households of all AIs
2010 2011 2012 2013
(%) H1 H2 H1 H2 H1 H2 H1
Mortgages 5.1 8.6 5.5 1.2 2.5 5.0 3.1Credit cards -0.9 17.9 -1.4 15.9 -1.6 15.3 -4.0Other loans for private purposes 7.9 6.6 9.4 3.8 5.0 9.3 10.9
Total loans to households 5.1 8.9 5.6 2.7 2.6 6.5 3.8
Source: HKMA.
The credit risk of unsecured household exposure
remained contained, with the annualised credit
card charge-off ratio and the number of
bankruptcy petitions staying low, albeit edging
up slightly (Chart 5.13).
Chart 5.13Charge-off ratio for credit card lending andbankruptcy petitions
Quarterly annualised charge-off ratio (%)Cases
Sources: Official Receiver’s Office and the HKMA.
0
1,000
2,000
3,000
4,000
5,000
6,000
0
1
2
3
4
5
2006 2007 2008 2009 2010 2011 2012 2013(Q1 - Q2)
Number of bankruptcy petitions during the quarter (lhs)Charge-off ratio (rhs)
38 Loans to households constitute lending to professional and private individuals, excluding lending for other business purposes. Mortgage lending accounts for a major proportion of household loans while the remainder comprises mainly unsecured lending through credit card lending and other personal loans for private purposes. At the end of June 2013, the share of household lending in domestic lending was 29%.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201366
banking sector performance
Meanwhile, the delinquency ratio of banks’
mortgage portfolios remained healthy
(Chart 5.14). However, it is worth noting that
the debt-service index of new mortgages further
deteriorated to 45 in June 2013, from 43 at the
end of 2012. The increase in debt-servicing
burden mainly reflected the rise in both the
average size of new mortgage loans and the
effective mortgage rate. Banks should be vigilant
about the impact of a rise in interest rates on
their mortgage portfolios, both in terms of
households’ debt repaying ability and the risk of
a possible property price correction.
Chart 5.14Delinquency ratio of banks’ mortgageportfolios and household debt-servicingburden in respect of new mortgages
1998 Q2 = 100%
Note: The calculation of the index is based on the average interest rate for BLR-based mortgages.
Sources: HKMA and staff estimates.
0
0.3
0.6
0.9
1.2
1.5
1.8
0
20
40
60
80
100
Mortgage delinquency and rescheduled loan ratio (lhs)Debt-service index of new mortgages (rhs)
1997 1999 2001 2003 2005 2007 2009 2011 2013(Q1 - Q2)
Corporate exposure 39
The growth of domestic loans to corporations 40
accelerated to 13.2% in the first half of 2013,
from 3.0% in the second half of 2012. At the end
of June 2013, corporate loans accounted for
70.5% of domestic lending.
There are some initial signs that the credit risk of
banks’ corporate exposures may be building up.
While the number of compulsory winding-up
orders of companies (Chart 5.15) and the
Altman’s Z-score 41 remained steady (Chart 5.16),
the debt leverage of the corporate sector has
increased in recent years, with the ratio of assets
to shareholders’ fund reaching 1.76 times at the
end of 2012 (Chart 5.17). Meanwhile, the
interest coverage ratio of local corporations,
which gauges their abilities to cover interest
expenses by earnings, showed a marked
deterioration. These indicators suggest that the
debt-servicing ability of the corporate sector
could be under test when interest rates rise.
Chart 5.15Number of winding-up orders of companies
Cases
Source: Official Receiver’s Office.
1998 2000 2002 2004 2006 2008 2010 2012 2013(H1)
0
100
200
300
400
500
600
700
800
Chart 5.16Altman’s Z-score: A bankruptcy riskindicator of listed non-financial companies
Z-score
Note: A lower Z-score indicates a higher likelihood of a company default.Source: HKMA staff estimates based on data from Bloomberg.
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
(H1)
0
1
2
3
4
5
6
7
75th Percentile
Median
25th Percentile
Increase in risk
39 Excluding interbank exposure.
40 Loans to corporations comprise domestic lending except lending to professional and private individuals.
41 Altman’s Z-score is a credit risk measure based on accounting data. It is a typical credit risk measure to assess the health of the corporate sector based on an array of financial ratios reported in companies’ financial statements. The accounting ratios used to derive the Z-score are working capital/total assets, retained earnings/total assets, earnings before interest and taxes/total assets, market value of equity/book value of total liabilities, and sales/total assets.
67 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Chart 5.17Leverage ratio and interest coverage ratio oflisted non-financial companies in Hong Kong
Notes:1. The leverage ratio is defined as the ratio of total assets to shareholders’ funds.
A higher value indicates higher leverage.2. Interest coverage ratio is defined as the ratio of earnings before interest and
taxes to interest expense.
Source: HKMA staff estimates based on data from Bloomberg.
%
0
14
130
200
Interest coverage ratio (lhs) Leverage ratio (rhs)
2
4
6
8
10
12
140
150
160
170
180
190
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
To the extent that a large outflow of funds from
Hong Kong may take place as a result of Fed
tapering, as some market participants anticipate,
the situation would be aggravated. Box 4
examines the factors behind the rising loan
spreads of syndicated loans after the global
financial crisis. The findings suggest that even if
the near-zero interest rate environment remains
unchanged in the near term, a significant
tightening of domestic liquidity conditions due
to external factors could drive up the loan
pricing in Hong Kong noticeably.
Mainland exposureThe credit exposure of the domestic banking
sector to Mainland-related business continued to
expand further. The total non-bank Mainland
exposure of all AIs increased to HK$3,198 billion
(18.6% of total assets) at the end of June 2013,
from HK$2,739 billion (16.2% of total assets) six
months earlier. Of this, retail banks’ total non-
bank Mainland exposure 42 rose to HK$2,003
billion (18.2% of total assets) from HK$1,779
billion (16.5% of total assets).
Chart 5.18Non-bank Mainland exposures of all AIs
% of total assetsHK$ bn
2010 2011 2012 2013(Q1 - Q2)
20092007 2008
Non-bank Mainland exposures (lhs)As a percentage of total assets of all AIs (rhs)
Note: Figures include exposures booked in AIs’ banking subsidiaries in Mainland China.
Source: HKMA.
400
800
1,200
1,600
2,000
2,400
3,600
3,200
2,800
0
2
4
6
8
10
12
14
16
20
18
0
The rising share of banks’ Mainland exposure
continues to be a significant risk factor amid
growing market concerns about Mainland
China’s growth outlook, corporate leverage and
funding conditions.
While banks’ lending to Mainland-related
customers is largely backed by guarantees or
collateralised, in view of the Mainland’s high
level of credit-to-GDP ratio (Chart 5.19), the
recent deterioration of the aggregate distance-to-
42 Including exposure booked in retail banks’ banking subsidiaries in Mainland China.
Chart 5.19Credit-to-GDP ratio in Mainland China
% of annual nominal GDP
100
110
120
130
140
150
200820072006 2009 2012 2013(Q1 - Q2)
20112010
Note: Credit-to-GDP ratio is defined as the ratio of claims on private sector to the sum of quarterly nominal GDP for the latest four quarters.
Sources: IMF International Financial Statistics and CEIC.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201368
banking sector performance
default index 43 of Mainland’s corporate sector
(Chart 5.20) and the rise in the amount of
non-performing loans in its banking system
(Chart 5.21), the need for Hong Kong banks to
maintain their stringent prudential management
of their Mainland exposure cannot be
over-emphasised.
Chart 5.20Distance-to-default index for the Mainlandcorporate sector
Index(Jun 2003 = 100)
Note: Distance-to-default index is defined as the simple average of the distance-to-default values of non-financial constituent companies (i.e. excluding investment companies and those engaged in banking, insurance and finance) of the Shanghai Stock Exchange 180 A-share index
Source: HKMA staff estimates.
Increase in risk
200820072006 2009 2010 20122011 2013(Jan - Aug)
0
20
40
60
80
100
Chart 5.21Non-performing loans in Mainland China
RMB bn
0.0
0.5
1.0
1.5
2.0
2.5
Source: China Banking Regulatory Commission.
Non-performing loans ratio (lhs)
Amount of non-performing loans (rhs)
2009 2010 2011 2012 2013(Q1 -Q2)
0
100
200
300
400
500
600
%
Impact of the European sovereign debt crisisWhile recent policies have reduced the tail risk of
the European sovereign debt crisis, downside
risks to economic growth remain. Thus, the
performance of local banks will continue to be
affected by the evolution of the European
sovereign debt crisis and fiscal issues. Given that
the exposure of the Hong Kong banking sector to
banks in the UK, France and Germany is not
immaterial (Chart 5.22), and these banks in turn
have significant exposures to the more
debt-ridden European economies, the possible
contagion risk and its implications for banks in
Hong Kong merit close attention.
Chart 5.22External claims of the Hong Kong bankingsector on major economies (all sectors) at theend of June 2013
Debt-ridden European economies Mainland China
Singapore UK
US Japan
France Germany
Others
Notes:1. Figures may not add up to 100% due to rounding.2. Debt-ridden European economies refer to Greece, Ireland, Italy, Portugal and
Spain.
Source: HKMA.
0.3%
36.5%
7.3%
6.2%6.3%4.8%
3.3%
1.7%
33.7%
43 The distance-to-default is a market-based default risk indicator based on the framework by R. Merton (1974), “On the pricing of corporate debt: the risk structure of interest rates”, Journal of Finance, Vol. 29, pages 449 - 470, in which equity prices, equity volatility, and companies’ financial liabilities are the determinants of default risk. In essence, it measures the difference between the asset value of a firm and a default threshold in terms of the firm’s asset volatility.
69 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Macro stress testing of credit risk 44 & 45
Results of the latest macro stress testing on retail
banks’ credit exposure suggest that the Hong
Kong banking sector remains resilient and should
be able to withstand rather severe
macroeconomic shocks, similar to those
experienced during the Asian financial crisis.
Chart 5.23 presents the simulated future credit
loss rate of retail banks in the second quarter of
2015 under four specific macroeconomic
shocks 46 using information up to the second
quarter of 2013. The expected credit losses for
retail banks’ aggregate loan portfolios two years
after the different hypothetical macroeconomic
shocks are estimated to be moderate, ranging
from 0.18% (interest rate shock) to 0.51% (Hong
Kong GDP shock).
Chart 5.23The mean and value-at-risk statistics ofsimulated credit loss distributions1
Mean to 90th percentile
90th percentile to 95th percentile
95th percentile to 99th percentile
99th percentile to 99.9th percentile
Notes: 1. The assessments assume the economic conditions in 2013 Q2 as the current
environment. The Monte Carlo simulation method is adopted to generate the credit loss distribution for each scenario.
2. Baseline scenario: no shock throughout the two-year period.3. Stressed scenarios: Hong Kong GDP shock: reductions in Hong Kong’s real GDP by 2.3%, 2.8%,
1.6%, and 1.5% respectively in each of the four consecutive quarters starting from 2013 Q3 to 2014 Q2.
Property price shock: Reductions in Hong Kong’s real property prices by 4.4%, 14.5%, 10.8%, and 16.9% respectively in each of the four consecutive quarters starting from 2013 Q3 to 2014 Q2.
Interest rate shock: A rise in real interest rates (HIBORs) by 300 basis points in the first quarter (i.e. 2013 Q3), followed by no change in the second and third quarters and another rise of 300 basis points in the fourth quarter (i.e. 2014 Q2).
Mainland GDP shock: Slowdown in the year-on-year annual real GDP growth rate to 4% in one year.
Source: HKMA staff estimates.
Mean
0 1 2 3 4 5 6
Credit loss as a % of the loan portfolio (%)
0.92
0.65
1.29
1.55
0.61
0.25
0.18
0.40
0.51
0.17
Mainland GDP shock
Interest rate shock
Property price shock
Hong Kong GDP shock
Stressed scenarios:3
Baseline scenario2
Taking account of tail risk, banks’ maximum
credit losses (at the confidence level of 99.9%)
under the stress scenarios range from 0.65%
(interest rate shock) to 1.55% (Hong Kong GDP
shock), which are significant, but smaller than
the loan loss of 4.39% following the Asian
financial crisis.
44 Macro stress testing refers to a range of techniques used to assess the vulnerability of a financial system to “exceptional but plausible” macroeconomic shocks. Details of the model adopted in this exercise can be found in J. Wong et al. (2006), “A framework for stress testing banks’ credit risk”, Journal of Risk Model Validation, Vol. 2(1), pages 3 - 23. An updated framework is used for the current estimations.
45 All estimates of credit loss for the overall loan portfolio of Hong Kong banks presented in this report are based on a revised stress testing framework. They are not strictly comparable to those estimates from the past framework that appeared in previous reports due mainly to different definitions of credit losses in these two frameworks. Specifically, credit losses in two years after any shock under the revised framework are measured by the estimated specific provision ratio at the end of the second year plus 50% of the estimated specific provision ratio at the end of the first year after the shock, while credit loss estimates from the past framework are derived based on an estimated delinquency ratio at the end of the second year multiplied by a loss-given-default estimate, which is determined by the simulated property price change over the two-year horizon.
46 These shocks are calibrated to be similar to those that occurred during the Asian financial crisis, except the Mainland China GDP shock.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201370
banking sector performance
5.5 Systemic risk to the banking system
While the credit default swap spreads for
European banks remained well above the levels
prevailing prior to the onset of the European
sovereign debt crisis, the corresponding spreads
for Asian banks continued to stay low
(Chart 5.24), suggesting the systemic risk of
banking sector in Asia has not been severely
affected by the sovereign debt crisis in Europe.
Chart 5.24Credit default swap spreads of banks in Europeand Asia
Basis points
2007 2008 2009 2013(Jan - Aug)
2011 20122010
Other European countries
Peripheral Europe Asia Pacific countries
Notes: 1. Median of five-year credit default swap spreads of the respective
groups.2. Peripheral Europe includes Greece, Ireland, Italy, Portugal and Spain.
Source: Bloomberg.
200
400
600
800
1,000
1,200
1,400
0
The Europeansovereign debt crisis
In Hong Kong, the banking distress index, a
market-based systemic risk indicator for the local
banking sector, increased slightly to 1.2 in
August from 1 in February 2013 (Chart 5.25).
However, the risk of contagion in the banking
system remained insignificant and the
probability of an occurrence of multiple bank
defaults is small. This optimistic market view was
broadly consistent with the latest assessment
result of the composite early warning system 47,
which estimated that the banking distress
probability remained within the range of the low
fragility category, suggesting that the banking
sector continued to be stable and resilient. 48
Chart 5.25The banking distress index
Index (Jan 1998 = 100)
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
The onsetof the global financialcrisis
Lower risk
SARSITbubble
2013
(Jan
-Aug
)
Asianfinancialcrisis
0
20
40
60
80
100
Source: HKMA staff estimates based on data from Bloomberg.
The proposed reform of the OTC derivatives
market being put forward by the Financial Stability
Board (FSB) is expected to play a significant role in
reducing the probability of a reoccurrence of
financial crises. The Macroeconomic Assessment
Group on Derivatives (MAGD) of the FSB
conducted an assessment of the costs and benefits
of the OTC derivatives reform. Box 5 presents the
key findings. The reform could reduce
counterparty risks through central clearing and
more comprehensive collateralisation. Such
benefits are balanced against the costs of holding
more capital and collateral by derivatives users.
MAGD quantitatively estimates that the reform
can generate median net benefits equivalent to
0.12% of GDP across 16 jurisdictions. For Hong
Kong, it is estimated that the net benefits are
around 0.13% of GDP.
Key performance indicators of the banking sector are
provided in Table 5.C.
47 The composite early warning system is designed to estimate banking distress probability based on 10 leading indicators. These include macroeconomic fundamentals, currency crisis vulnerability, default risk of banks and non-financial companies, asset price misalignments, credit growth, and the occurrence of banking distress in other Asia-Pacific economies. For details, see J. Wong et al. (2010), “Predicting banking distress in the EMEAP economies”, Journal of Financial Stability, Vol. 6(3), pages 169 - 179.
48 The composite early warning system is a four-level risk rating system. A. Demirgüc-Kunt and E. Detragiache (2000), “Monitoring Banking Sector Fragility: A Multivariate Logit Approach”, World Bank Economic Review, Vol. 14(2), pages 287 - 307, has been followed in the selection of the upper bounds of each of the four fragility classes so that type I error associated with the bounds are 10%, 30%, 50% and 100% respectively.
71 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Table 5.CKey performance indicators of the banking sector1 (%)
Jun 2012 Mar 2013 Jun 2013
interest rate1-month HIBOR fixing2 (quarterly average) 0.30 0.23 0.213-month HIBOR fixing (quarterly average) 0.40 0.39 0.38BLR3 and 1-month HIBOR fixing spread (quarterly average) 4.70 4.77 4.79 BLR and 3-month HIBOR fixing spread (quarterly average) 4.60 4.61 4.62Composite interest rate4 0.42 0.25 0.32
Retail banks
balance sheet developments 5
Total deposits 1.3 -0.3 1.4Hong Kong dollar 0.2 -1.6 0.5Foreign currency 2.7 1.5 2.5
Total loans 3.4 2.2 5.6Domestic Lending 6 2.9 2.5 5.3 Loans for use outside Hong Kong 7 6.1 0.9 7.1
Negotiable instrumentsNegotiable certificates of deposit (NCD) issued 6.4 2.8 15.1Negotiable debt instruments held (excluding NCD) -2.8 4.5 -0.3
asset quality 8
As a percentage of total loansPass loans 98.19 98.32 98.44Special mention loans 1.28 1.22 1.11Classified loans 9 (gross) 0.52 0.46 0.45Classified loans (net) 10 0.30 0.32 0.31Overdue > 3 months and rescheduled loans 0.45 0.35 0.34
profitabilityBad debt charge as percentage of average total assets 11 0.01 0.02 0.03Net interest margin 11 1.35 1.39 1.41Cost-to-income ratio 12 44.5 r 41.7 40.9
liquidity ratio (quarterly average) 39.7 40.2 38.9
Surveyed institutions
asset qualityDelinquency ratio of residential mortgage loans 0.01 0.01 0.02Credit card lending
Delinquency ratio 0.21 0.23 0.25Charge-off ratio — quarterly annualised 1.95 r 1.81 2.06
— year-to-date annualised 1.74 r 1.81 1.88
All locally incorporated AIs
capital adequacy ratio (consolidated) 13 15.9 16.6 15.9
Notes:
1. Figures are related to Hong Kong office(s) only except where otherwise stated.2. The Hong Kong dollar Interest Settlement Rates are released by the Hong Kong Association of Banks. 3. With reference to the rate quoted by The Hongkong and Shanghai Banking Corporation Limited.4. The composite interest rate is a weighted average interest rate of all Hong Kong dollar interest-bearing liabilities, which include deposits from customers,
amounts due to banks, negotiable certificates of deposit and other debt instruments, and Hong Kong dollar non-interest-bearing demand deposits on the books of banks. Further details can be found in the HKMA website.
5. Quarterly change.6. Loans for use in Hong Kong plus trade finance.7. Including “others” (i.e. unallocated).8. Figures are related to retail banks’ Hong Kong office(s) and overseas branches.9. Classified loans are those loans graded as “substandard”, “doubtful” or “loss”.10. Net of specific provisions/individual impairment allowances.11. Year-to-date annualised.12. Year-to-date figures.13. For the implementation of the Basel III framework, all locally incorporated AIs are required to report their capital adequacy positions under the Banking
(Capital) Rules starting from 1 January 2013.r Revised figure.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201372
banking sector performance
Box 4Explaining rising loan spreads on syndicated loans in
Hong Kong after the global financial crisis
IntroductionSpreads on syndicated loans in Hong Kong rose
significantly after the global financial crisis from
an average of below 100 basis points over HIBOR
or LIBOR to around 250 basis points
(Chart B4.1) 49. This box empirically identifies
factors contributing to the rising loan spread and
assesses the implications.
Chart B4.1Average spreads on syndicated loans inHong Kong
Note: Loan spreads are expressed as the four-quarter moving average.
Sources: Bloomberg and HKMA staff estimates.
Average basis points above HIBOR or LIBOR
2006 2007 2009 2011 20122008 2010 2013(Q1 - Q2)
0
300
US dollar loans Hong Kong dollar loans
50
100
150
200
250
Possible explanations for the rising spread There are some factors that could account for the
rising loan spread on syndicated loans in Hong
Kong after the global financial crisis. First, the
higher loan spread may reflect a riskier
composition of borrowers tapping syndicated
loans in Hong Kong. This may be associated
with the rapid growth of syndicated loans to
Mainland-related firms. Besides, borrowers with
unrated or speculative-grade ratings are observed
to be more prevalent.
Second, the rising loan spread may be partly a
result of higher default risk premiums demanded
by banks to compensate for the potential adverse
impact of weak business conditions after the
global financial crisis on firms’ financial health.
In fact, the spike in the loan spread that occurred
from the fourth quarter of 2008 to the third
quarter of 2009 is associated with a pessimistic
business outlook (Chart B4.2).
Chart B4.2Expected changes of Hong Kong businesssituation1 and spreads on syndicated loans
%Average basis points above HIBOR or LIBOR
0
300
-60
-40
-20
0
20
40
Notes:1. An increase in the expected change of business situation indicates that
business situation is likely to improve, while a decrease may signal deterioration in business situation in the near term.
2. Loan spreads are expressed as the four-quarter moving average.
Sources: Bloomberg, C&SD and HKMA staff estimates.
Expected changes of business situation (rhs)
US dollar loan spreads (lhs)
Hong Kong dollar loan spreads (lhs)
50
100
150
200
250
2007 20082005 2006 2009 2010 2011 2012 2013(Q1 - Q2)
The third factor is that banks may charge higher
liquidity premiums for syndicated loans after the
global financial crisis. This hypothesis is
supported by the fact that the foreign-currency
loan-to-deposit (LTD) ratio of the Hong Kong
banking sector has shown a clear upward trend
since 2009 (see Chart 5.8 in Chapter 5), which
may signal significant pressure on banks’
funding, particularly for their US dollar lending.
49 See He, D. and R. N. McCauley (2013), “Transmitting Global Liquidity to East Asia: Policy rates, Bond Yields, Currencies and Dollar Credit”, paper prepared for presentation at the Hong Kong Monetary Authority 20th Anniversary Research Workshop, 7 June 2013, Hong Kong.
73 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Finally, short-term interest rates may be one
factor affecting loan spreads, although the
direction of the impact may be inconclusive. On
one hand, a negative relationship is consistent
with Merton’s (1974) theory of corporate
default. 50 On the other hand, it may be argued
that firms may take advantage of lower
short-term interest rates to refinance debt, thus
reducing their debt-servicing burdens and default
risk. This counter argument implies a positive
relationship between short-term interest rates
and loan spreads.
An empirical model of loan spreadsTo what extent these four factors account for the
rising loan spread can be studied empirically.
Here, a two-stage approach is adopted to
decompose the relative contribution of each
factor. In the first-stage analysis, we investigate
how risk characteristics of individual borrowers
help explain the loan spread. Specifically, using
a database of 737 loan observations from the first
quarter of 2005 to the second quarter of 2013, we
regress the loan spread by a set of conventional
loan-level default risk indicators.
The estimation result (Table B4.A) suggests that
syndicated loans with smaller loan sizes, longer
maturities and speculative grade ratings tend to
be associated with higher loan spreads. 51
Mainland-related loans are estimated with a
higher loan spread, probably reflecting the strong
demand for foreign-currency loans by Mainland
firms such that they are more willing to pay a
premium for syndicated loans in Hong Kong. In
addition, banks are estimated to charge a lower
loan spread for Hong Kong dollar loans than US
dollar loans. Largely in line with findings in the
literature 52, loans backed by collateral are found
to be associated with higher loan spreads, which
could be explained partly by the banks being
more likely to demand collateral for loans that
are perceived to entail high default risk. As
revealed from the R2, a conventional goodness-
of-fit statistic, risk characteristics of individual
borrowers only account for around 17% of the
variation of loan spreads, suggesting that macro
factors (i.e. the remaining three factors) may play
more significant roles in determining loan
spreads on syndicated loans in Hong Kong.
Table B4.AEstimation result for the first-stage regression
explanatory variables
dependent variable:Spreads on syndicated loans
(Sample period:2005Q1 - 2013Q2)
Log(Size of loan facility) -5.22 *Time to maturity1 1.35 ***A dummy variable for loans with speculative
grade ratings2 84.79 ***A dummy variable for Mainland-related loans 36.27 ***A dummy variable for Hong Kong dollar loans -37.22 ***A dummy variable for loans with collateral 32.96 ***Constant 232.78 ***R2 0.171Number of observations 737
Notes:
1. In number of months, for loans with maturity shorter than 60 months.
2. Loans with ratings below BBB.
3. *** and * denote significance at 1% and 10% levels respectively.
4. The estimation sample includes Hong Kong dollar loans and US dollar loans.
50 See Merton, R. (1974), “On the Pricing of Corporate Debt: The Risk Structure of Interest Rates”, Journal of Finance, 29, pages 449 – 470. Specifically, an important part of the theory is that a firm’s expected growth rate of asset is determined by risk-free interest rates. Assuming a constant amount of debt, a decline in interest rates leads to a reduction in the firm’s net worth, implying higher default risk and thus a higher loan spread. While empirical evidence remains mixed, Lo and Hui (2013) recently find a significant negative relationship between credit spreads and risk-free interest rates in the US during January 2008 to June 2013. For details, see Lo and Hui (2013), “Pricing Corporate Bonds in an Ultra-low Interest Rate Environment”, available at the Social Science Research Network (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2185861).
51 A preliminary analysis finds that a dummy variable for guarantee and that for loans with investment grade ratings are statistically insignificant. These two variables are therefore dropped from the regression equation.
52 See Jiménez, G., and J. Saurina (2004), “Collateral, Type of Lender and Relationship Banking as Determinants of Credit Risk”, Journal of Banking and Finance, 28(9), pages 2191 – 2212.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201374
banking sector performance
The second-stage analysis focuses on the three
macro factors. Specifically, we first compute the
residuals from the first-stage regression equation
(denoted by e), which by construction contain
information other than risk characteristics of
individual borrowers. We then regress e by an
indicator of expected changes in business
situation (EXP) 53 , LTD ratios, an interacting term
between a dummy variable 54 and LTD ratios
(LTD09), and short-term interest rates (SR).
LTD09 is included to capture a possible increase
in the sensitivity of the loan spread to LTD ratios
when LTD ratios are on their rising trends. Since
these macro factors may have different impacts
on Hong Kong dollar loans and US dollar loans,
we estimate the equation separately for these two
groups of samples.
In line with our discussion in the previous
section, we hypothesise that the loan spread is
correlated negatively with EXP and positively
with the LTD ratio. For LTD09, we expect a
positive estimated coefficient for two reasons.
First, there is a clear rising trend of the foreign-
currency LTD ratio since 2009, which may signal
significant funding pressure on banks. Second,
since late 2009, the stance of macroprudential
policy in Hong Kong has been tightened
generally, partly reflecting regulatory concerns
about the sustainability of credit growth and
rising LTD ratios. 55 Banks may since then
become more prudent in pricing their lending,
including syndicated loans, by factoring in the
potential liquidity risk amid the unprecedented
low interest rate environment.
The estimation result (Table B4.B) is generally in
line with our expectation, except for short-term
interest rates, which is found to be statistically
insignificant. 56 Nevertheless, the macro factors
together account for around 32% and 18% of the
variation of loan spreads for Hong Kong dollar
loans and US dollar loans respectively 57,
suggesting that macroeconomic environments
significantly affect the pricing for syndicated
loans in Hong Kong.
Table B4.BEstimation result for the second-stage regression
dependent variable: e(Sample period: 2005Q1 - 2013Q2)
explanatory variables HKd loans USd loans
SR -4.02 -4.94EXP -1.90 *** -1.32 ***LTD ratio 1.79 ** 0.34LTD09 1.13 *** 1.17 **Constant -148.23 ** -20.83R2 0.385 0.218Number of observations 353 384
Note: *** and ** denote significance at 1% and 5% levels respectively.
53 An increase in EXP indicates that business situation is likely to improve while a decrease may signal deterioration in business situation in the near term.
54 The dummy variable is defined as one after November 2009 for Hong Kong dollar loans and after July 2009 for US dollar loans.
55 Apart from a series of macroprudential measures since late 2009 to help banks to manage specifically their risks in mortgage lending, the HKMA issued circular “Credit growth” on 11 April 2011 to all AIs, requiring them to reassess their loan business and funding plans.
56 Short-term interest rates are found to have a significant negative relationship with the loan spread in univariate regression analysis. However, the significance does not carry over after controlling for the effect of LTD ratios.
57 This is calculated as where and are the R-squared statistics from the first- and second-stage estimations respectively.
75 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Decomposition analysis of loan spreads Based on the estimation result, we decompose
the relative contribution of each factor to
changes in the loan spreads. Charts B4.3 and
B4.4 show the cumulative change in the loan
spreads since the third quarter of 2007 for Hong
Kong dollar loans and US dollar loans
respectively together with the estimated
contribution by each factor. There are some
interesting patterns from the charts. First, risk
characteristics of individual borrowers are found
to contribute mainly to the volatility of the loan
spread rather than the upward trend (see the red
bars). Second, among the macro factors,
liquidity, which is proxied by LTD ratios,
contributes the most significant part of the rising
loan spread (i.e. the orange bars), followed by
changes in expectation of business situation
(i.e. the blue bars). Changes in short-term
interest rates, even assuming it as one significant
contributor, are found to have a limited
contribution to the rising loan spread.
Chart B4.3Factors contributing to the rising loan spreadon Hong Kong dollar syndicated loans since2007Q3
bps
-50
300
Note: Changes in loan spreads and factor contributions in each period are the weighted averages by loan size.
Sources: Bloomberg, C&SD and HKMA staff estimates.
0
50
100
150
200
250
2008 2009 2010 2011 2012 2013(Q1 - Q2)
2007(Q3 - Q4)
LTD ratio
Expected changes of business situation in Hong Kong
Short-term interest rate
Risk characteristics of individual borrowers
Changes in loan spreads since 2007Q3
Chart B4.4Factors contributing to the rising loan spreadon US dollar syndicated loans in Hong Kongsince 2007Q3
bps
2008 2009 2010 2011 2012 2013(Q1 - Q2)
2007(Q3 - Q4)
-100
-50
300
Note: Changes in loan spreads and factor contributions in each period are the weighted averages by loan size.
Sources: Bloomberg, C&SD and HKMA staff estimates.
0
50
100
150
200
250
LTD ratio
Expected changes of business situation in Hong Kong
Short-term interest rate
Risk characteristics of individual borrowers
Changes in loan spreads since 2007Q3
ConclusionThis analysis empirically reveals that less
abundant domestic liquidity conditions, as
reflected by rising LTD ratios since 2009, are one
major contributor to the rising loan spread of
syndicated loans in Hong Kong. One implication
is that even if the near-zero interest rate
environment remains unchanged in the near
term, a significant change in domestic liquidity
conditions due to external factors could drive the
loan pricing in Hong Kong noticeably. There is
market speculation that the US tapering could
result in significant outflows of funds from Asia,
including Hong Kong. If such a scenario occurs,
domestic liquidity conditions could be tightened
due to withdrawals of deposits, and loan prices
could rise significantly ahead of any interest rate
hike. The potential impact of a sudden change
in local liquidity conditions on loan prices and
its possible consequences merit close attention.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201376
banking sector performance
Box 5Macroeconomic impact assessment of the OTC derivatives reform
Background The over-the-counter (OTC) derivatives market
expanded rapidly in the past decade, with the
outstanding amount of derivative contracts rising
from US$80 trillion at the end of 1998, to
US$633 trillion at the end of 2012 (Chart B5.1).
It is generally believed that the opaqueness,
highly unregulated and insufficiently
collateralised nature of OTC derivatives market,
given its sheer size, helped propagate and
amplify the global financial crisis. In response,
the G20 leaders declared in the 2009 Pittsburgh
Summit that more regulations on OTC
derivatives were required. At the initiative of the
Financial Stability Board (FSB), a working group
was formed in April 2010 to make
recommendations on the implementation of the
G20 decisions. 58 Specifically, the proposed OTC
derivatives reform consists of four elements:
(1) all standardised OTC derivative contracts
should be traded on exchanges or electronic
trading platforms, where appropriate, and cleared
through central counterparties (CCPs); (2) OTC
derivative contracts should be reported to trade
repositories; (3) non-centrally cleared contracts
should be subject to higher capital requirements;
and (4) non-centrally cleared contracts should be
subject to more stringent margin requirements 59 .
Chart B5.1Outstanding amount of global OTC derivativescontracts
Note: Figures are based on Semi-annual Over-the-counter (OTC) Derivatives Markets Statistics published by the BIS.
Source: BIS.
0
100
200
300
400
500
600
700
800
1998 2000 2002 2004 2006 2008 2010 2012
UnallocatedCredit default swapsCommodity contractsEquity-linked contractsInterest rate contractsForeign exchange contracts
US$ trillions
As the reform is still under consultation, the
Over-the-Counter Derivatives Coordination
Group commissioned the Macroeconomic
Assessment Group on Derivatives (MAGD) to
conduct a quantitative benefits and costs
assessment of the macroeconomic implications
of the reform. At the invitation of the Bank for
International Settlements (BIS), the HKMA has
joined the MAGD. 60 This box presents key
findings of the assessment.
58 For a progress report of the reform across member institutions of the FSB, see Sixth Progress Report on Implementation of OTC Derivatives Market Reforms, published by the FSB in September 2013. In Hong Kong, to provide for the regulatory framework for the OTC derivatives market, the relevant Bill was gazetted in June 2013 and was tabled before the Legislative Council in July 2013. Detailed rules for implementing the new framework, in the form of subsidiary legislation, are being prepared and will be issued for public consultation in late 2013.
59 Initial margins and variation margins are deployed in OTC derivatives transactions to mitigate counterparty risks. Initial margins are posted by each counterparty at the beginning of the trade to cover potential losses in the event of default. Variation margins are adjusted between counterparties daily to reflect mark-to-market profit and loss in their derivative positions.
60 MAGD is composed of 29 member institutions of the FSB, with the help of the BIS and International Monetary Fund. The full MAGD report, “Macroeconomic impact assessment of OTC derivative regulatory reforms”, published on 26 August 2013 is available in the BIS website.
77 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
Macroeconomic assessment
(A) Benefits of the reformBenefits of the reform are gauged by network
modelling. Based on data from the 16 largest
derivatives dealers (G16 dealers) and 25 less
active banks at the end of 2012, a network of
bilateral exposures between them is estimated. 61
The estimated network attempts to mimic the
structure of the OTC derivatives market, which is
characterised by a highly interconnected core
and less interconnected periphery (Chart B5.2). 62
Chart B5.2Stylised core-periphery network
P3
P2
P4P5
P1
C4 C3
C1 C2
Notes:
1. Orange nodes with a letter “C” refer to ‘core’ set of players in the network, and blue nodes with a letter “P” refer to ‘periphery’ set of players. The arrows represent the direction of bilateral gross exposure. For example, the arrow originates from P1 to C1 denotes a claim of P1 on C1.
2. Banks from the periphery set do not have linkages among themselves. However, they are indirectly linked to each others through the core banks.
To understand the propagation mechanism of the
estimated network, an exogenous shock is
considered to the G16 dealers that increases their
default probability (Chart B5.3). Counterparties
which have direct exposures to G16 dealers will
respond to the shock by increasing the credit
valuation adjustments (CVA) for the
uncollateralised derivatives exposures. 63
Although actual default is yet to occur, this results
in mark-to-market losses equal to changes in
CVAs and drives up counterparties’ leverage. Due
to a higher leverage, the riskiness of the network
increases and the losses reverberate further to
other banks that may not have direct exposures
to the G16 dealers initially. Overall, the shock to
the G16 dealers would lead to an increase in the
system-wide leverage. If the increase in leverage
is severe enough, the elevated systemic risk can
result in the occurrence of a financial crisis.
Chart B5.3Stylised model architecture
Counterpartyraises CVA
Mark-to-marketlosses
Changes incounterparty creditrisk that have direct
exposure to major dealers
increase in leverage
Increase system-wide leverage
Shock to majordealers
Increase probability offinancial crisis
There are two major benefits of the reform that
help banks to withstand shocks. First, as central
clearing through CCPs enables multilateral
netting, the cascade of losses will be transmitted
in a less complex network, thus the systemic risk
arising from inter-linkage between banks will be
significantly reduced. 64 Secondly, better
collateralisation against OTC derivatives
exposures can lower the impact of the initial
shocks when it reverberates through the
network. 65 The MAGD estimates that the reform
can lower the probability of occurrence of a
financial crisis in any given year by 0.26
percentage points. Based on the finding that the
61 As data for banks’ bilateral exposures are not available due to confidentiality issues, the network of bilateral exposures is estimated using an improved maximum entropy method. For details, see G. Halaj and C. Kok (2013), “Assessing Interbank Contagion Using Simulated Networks”, ECB Working Paper No. 1506.
62 To ensure the estimated network is a reasonable description of the reality, summary statistics generated from the estimated network have been validated against confidential data sources.
63 CVA is an adjustment made by the institution to the valuation of a netting set with the counterparty to reflect the market value of the credit risk of that counterparty. CVA capital charge represents the amount of capital that a bank is required to hold for the CVA risk of the counterparty and is a new element under the Basel III regulatory framework.
64 Netting is the process of offsetting different counterparties’ transactions into one single net obligation.
65 Under the Basel III capital requirement, a better collateralised exposure is subjected to a lower CVA charge.
Half-YearlY MonetarY and financial StabilitY report SepteMber 201378
banking sector performance
median cost of a financial crisis is 60% of the
pre-crisis GDP, it is estimated that expected
annual GDP losses of about 0.16% could be
avoided by the reform. 66 It is noteworthy that
the estimated benefits are based on data from the
derivatives exposure of banks which have
significant operations in the global OTC
derivatives market, thus the estimated benefits
are equally applicable to every jurisdiction.
(B) Costs of the reform There are three ways in which the reform may lead to higher cost to banks. First, the higher CVA capital charges require banks to hold more capital, which creates downward pressure on their profitability as cost of equity is typically greater than debt. 67 Second, the more stringent margins requirements for both centrally and non-centrally cleared OTC derivatives require banks to hold more collateral-eligible assets, which results in additional expenses. 68 Third, the fee required by
66 The benefits of the reform are computed as the reduction in the probability of the crisis times its output cost. The output cost is estimated from a previous macroeconomic assessment by the Macroeconomic Assessment Group. For details, see Basel Committee on Banking Supervision (2010), “An assessment of the long-term economic impact of stronger capital and liquidity requirements”.
67 It is noteworthy that the increase in funding cost is an upper-bound estimate since it ignores the contribution of a lower risk arising from a better capitalised banking system.
68 The cost increase in OTC derivatives due to additional collateral requirements is also an upper-bound estimate since it ignores the possible situation of a more favourable pricing arising from lower counterparty risk when counterparties post more collateral.
69 Default funds are made up of contributions from both clearing participants and CCPs. It has a risk-sharing feature that non-defaulting clearing participants may be required to share any losses due to a default of another clearing participant.
70 It is assumed that multilateral netting associated with central clearing is about four times more effective than bilateral netting. In designing various netting scenarios, the percentage of OTC derivatives that will be centrally traded is assumed to change from the current 35% to 60% post reform in the central scenario. In the high-netting scenario, the ratio is increased further to 75% post reform. In the low-netting case, while 75% of trades are assumed to clear through CCPs, there are no netting benefits.
71 Hong Kong’s interest rates and lending spreads are proxied by HIBOR and the net interest margins of the Hong Kong banking sector. For details, see Wong et al., “An Assessment of The Long-term Economic Impact of the New Regulatory reform on Hong Kong” HKMA Research Note 05/2010.
CCPs and the default fund contributions are direct expenses incurred by banks when more trades are shifted to central clearing. 69
To access the impact, based on the differences in the assumed effectiveness of the multilateral netting achieved by CCPs, three scenarios (high-netting, low-netting and central) are considered. 70 The MAGD estimates that the cost under the central scenario is €20 billion, while the costs under the high and low netting scenarios are €32 billion and €15 billion respectively. It is assumed that banks will transfer the costs to customers by widening the lending spreads. Using an estimate of €24 trillion that represents the size of the global banking loan book, the lending spreads are estimated to increase by 6, 8 and 13 basis points in the high netting, central and low netting scenarios respectively.
Using a suite of macroeconomic models provided by the MAGD members, it is estimated that the increase in lending spreads would reduce GDP by 0.03% (high-netting) to 0.07% (low-netting) under the three scenarios (Table B5.A). For Hong Kong, an error correction model is developed by linking Hong Kong’s GDP with its interest rates and lending spreads. 71 Under the three scenarios, it is estimated that the reform would lead to a drop of GDP ranging from 0.02% (high-netting) to 0.05% (low-netting).
Table B5.AMacroeconomic benefits and costs of OTC derivatives regulatory reforms
impact on long-run Gdp in per cent
High-netting scenario
central scenario
low-netting scenario
Global averageBenefits1 0.16 0.16 0.16Costs2 -0.03 -0.04 -0.07Net benefits 0.13 0.12 0.09
Hong KongBenefits1 0.16 0.16 0.16Costs3 -0.02 -0.03 -0.05Net benefits 0.14 0.13 0.11
Notes:
1. Calculated as the impact on GDP due to a financial crisis times the reduction of probability of a crisis. The assumed decline in output is 60% of pre-crisis GDP. As exposures were found to be sufficiently collateralised post reform, the reductions of probability of a crisis and hence the benefits are estimated to be the same under the three scenarios. The estimated benefits are equally applicable to every jurisdiction.
2. Impact on GDP due to higher lending spreads. The increases in lending spreads under the high-netting, central and low-netting scenarios are 6, 8 and 13 basis points respectively. The figures are based on the median of the results from 16 jurisdictions using different macroeconomic models.
3. Estimates are for Hong Kong only.
79 Half-YearlY MonetarY and financial StabilitY report SepteMber 2013
banking sector performance
(C) Net Result In sum, the MAGD concludes that the net
benefits of the reform are positive in all
scenarios, with a central estimate of 0.12% of
GDP a year. For Hong Kong, the net benefits of
the reform are about 0.13% of GDP a year.
Limitations of the quantitative assessmentAlthough the analytical approach suggests that
the reform can generate net benefits, there are a
number of factors that may affect the impact of
the reform, but their effects have not been
covered by the assessment. These include the
following:
1. Interconnection risks arising from multilateral
netting across CCPs
Central clearing allows CCPs to perform
multilateral netting of exposures, thereby
facilitating the reduction of counterparty
risk. Given the proliferation of CCPs, this
benefit can only be realised if there are
linkages among them. While such linkages
among CCPs could increase the scope for
multilateral netting and reduce collateral
demand, it may also introduce
interconnection risks and transmit
participants’ failure among themselves. The
potential impact of the establishment of
linkages among CCPs needs to be further
assessed.
72 For large institutions, it seems unlikely that they will be discouraged from using derivatives in hedging at all just because of the higher costs. Nonetheless, during a meeting organised by MAGD with OTC derivatives market participants, it was suggested that some smaller pension funds and non-financial companies may choose to hedge much less or not at all.
73 One example is the difference in the recognition of CCPs. For instance, European and US banks may be prevented from clearing their trades through CCPs that are not recognised by or registered in their home jurisdictions.
2. Pressures on prices of collateralised assets
It is foreseeable that collateral demand will
increase significantly post reform. Any
shortage of collateral during times of stress
may pose excessive pressures on prices of
high quality assets, and hence increase the
costs of using derivatives.
3. Increased cost of indirect clearing
Indirect clearing through other CCPs
clearing members offers a way through
which smaller market participants could
have access to CCPs. However, they may
face higher margin requirements required
by the direct clearing members than that
imposed by CCPs on direct clearers
themselves. The more stringent margin
requirements, which aim at protecting
direct clearing members from the increased
risk associated with indirect clearing, will
raise the cost of risk transfer. Should this
deter risk hedging by smaller market
participants, it may not be conducive to
economic and financial stability. 72
4. Cross-border issues and its impact
As jurisdictions around the world gradually
put their own regulatory regimes in place,
there are indications of potential differences
in the scope and application of the new
regulations across jurisdictions. 73 Should
any conflicts in the regulatory frameworks
not be properly addressed, there is a risk
that they could create potential for
regulatory arbitrage. Given the significant
presence of global dealers in the regional
OTC derivatives markets, any withdrawal of
them in taking advantage of the regulatory
differences may result in market
fragmentation, and may have a significant
impact on the pricing and liquidity of the
products traded in the region.